Assessing congestion in emergency obstetric care in public hospitals in Tamil Nadu

The purpose of this paper is to assess congestion as the simultaneous occurrence of desirable health output (e.g. maternal admissions) along with undesirable output (e.g. still births (SB)), in emergency obstetric care settings of public hospitals in Tamil Nadu.

The study is based on a cross-sectional data set of 97 public hospitals collected by the statistical cell of Tamil Nadu Health Systems Project for the year 2013–2014. The study uses three inputs – beds, doctors and nurses; three desirable outputs – maternal admissions, neonatal admission and live births; and four undesirable outputs – SB, intra-uterine deaths, neonatal deaths and maternal deaths. Congestion analysis, a variant of the data envelopment analysis (DEA) method and slack analysis, has been applied to detect an excessive use of some inputs or a shortfall in some outputs across these hospitals. Furthermore, the association between congestion and some contextual factors has been examined.

On an average, the hospitals in our sample can increase the total amount of outputs by 62.8 percent by improving overall efficiency, and about 34.2 percent of this inefficiency can be attributed to congestion. Analysis of sub-samples showed that government hospitals at the taluk level have higher congestion than district headquarter hospitals. Congestion seems to decrease with greater hospital volume up to a limit; beyond that, it increases in obstetric care settings.

Hospital-based efficiency studies in the Indian context, so far, have estimated relative efficiency among hospitals using the classical DEA method, but ignoring adverse health outcomes. Congestion analysis, an advance in the DEA method, considers how much the desirable outputs can be increased as also how much undesirable outputs affect efficiency.